Antibacterial Drug Shortages From 2001 to 2013: Implications for Clinical Practice

M A J O R A R T I C L E

Antibacterial Drug Shortages From 2001 to 2013:

Implications for Clinical Practice

Farha Quadri,1Maryann Mazer-Amirshahi,2Erin R. Fox,3Kristy L. Hawley,4Jesse M. Pines,1,5Mark S. Zocchi,4and Larissa May1

1_{Department of Emergency Medicine, George Washington University, and}2_{MedStar Washington Hospital Center, Washington D.C.;}3_{Drug Information} Service, University of Utah Health Care, Salt Lake City;4_{Office for Clinical Practice Innovation, and}5_{Department of Health Policy, George Washington} University, Washington D.C.

Background. Previous studies have described drug shortages; however, there has been no comprehensive evaluation focusing on US antibacterial shortages.

Methods. Drug shortage data from the University of Utah Drug Information Service database were analyzed, with a focus on antibacterial agents from 2001 to 2013. We used descriptive statistics to describe trends in drug shortages, analyze drug classes commonly affected, and investigate whether drugs experienced multiple periods of shortages.

Results. One hundred forty-eight antibacterial drugs were on shortage over the 13-year study period, with 26 drugs still active on shortage as of December 2013. The median number of new shortages per year was 10 (interquartile range [IQR], 7). The number of drugs on shortage increased at a rate of 0.35 additional drugs every month (95% confidence interval, .22–.49) from July 2007 to December 2013 (P< .001). The median shortage duration was 188 days (IQR, 366.5). Twenty-two percent of drugs experienced multiple shortage periods.

Conclusions. There were a substantial number of drug shortages from 2001 to 2013, with a dramatic rise in short-ages since 2007. Shortshort-ages of agents used to treat multidrug-resistant infections are of concern due to continued trans-mission and limited treatment options.

Keywords. drug shortages; University of Utah Drug Information Service (UUDIS); National Drug Codes (NDC).

Drug shortages are an important issue in public health and medical care delivery [1]. The US Food and Drug Administration (FDA) defines a drug shortage as “a situation in which the total supply of all clinically inter-changeable versions of an FDA-regulated drug product is inadequate to meet the projected demand at the user level”[2]. There are several reasons for drug shortages, including manufacturer mergers, facility consolidation, manufacturing quality issues, and narrow profit margin for generic drugs [3].

Anti-infective agents account for 15% of all drug short-ages, with injectable antibiotics frequently impacted [4]. A majority of infectious disease physicians surveyed

report having modified antimicrobial selection due to shortages [5]. Shortages have significant implications

in the context of other factors that limit availability of effective antimicrobials, including a shrinking pipeline of new antibiotics and increasing drug resistance [6– 8]. In this study, we evaluate antibacterial drug shortage trends in the United States from 2001 to 2013 and dis-cuss clinical implications. A recent government report highlighted the public health crisis of drug shortages, but to our knowledge, ours is thefirst in-depth evalua-tion of antibacterial drug shortages [9].

METHODS

We conducted a cross-sectional study of antibacterial drug shortages using the University of Utah Drug Infor-mation Service (UUDIS) database from January 2001 to December 2013. UUDIS began collecting national drug shortage data in January 2001, and it publishes critical drug shortage information on a public website (www. ashp.org/shortages) hosted by the American Society of Received 13 January 2015; accepted 5 March 2015.

Correspondence: Larissa May, MD, MSPH, MSHS, The George Washington University, Department of Emergency Medicine, 2120 L St NW, Ste 450, Washington D.C. 20037 ([email protected]).

Clinical Infectious Diseases®

© The Author 2015. Published by Oxford University Press on behalf of the Infectious Diseases Society of America. All rights reserved. For Permissions, please e-mail: [email protected].

DOI: 10.1093/cid/civ201

Health-System Pharmacists (ASHP). The UUDIS defines a shortage as a supply issue affecting how the pharmacy prepares or dispenses a drug product or influences patient care when pre-scribers must use an alternative agent [10]. For this study’s pur-pose, we used this definition for a shortage, which is more inclusive than the FDA definition. Detailed UUDIS methods have been previously published [11]. UUDIS receives voluntary drug shortage reports via the reporting feature on the ASHP website. Clinical pharmacists at UUDIS verify the existence of a shortage and determine all potential manufacturers of a drug reported to be in short supply using National Drug Codes (NDCs). Next, each manufacturer is contacted to deter-mine which NDCs are in shortage at the national level. The manufacturers are also asked for reasons for the shortage and an estimated release date. UUDIS posts information at the ASHP drug shortage website, noting which products are affect-ed, methods for accessing available products, reasons for the shortage, and estimated resupply dates. If applicable, patient care implications, safety concerns, therapeutic alternatives, and management strategies are also provided. UUDIS considers a shortage to be resolved when all suppliers have all NDCs (available strength and product size) available or have discon-tinued their products. UUDIS collects the following drug short-age data using Microsoft Excel: generic product name, therapeutic category, date shortage began (date UUDIS was no-tified), resolved date, duration, reason for shortage, and if the drug is an injectable product. Drug shortages that occur due to product discontinuation or withdrawal from the market are also included in the UUDIS database and marked as having shortage duration of zero days [10].

We analyzed drug shortage data from UUDIS from January 2001 to December 2013. Three investigators evaluated 1751 pharmaceutical products with reported shortages, focusing on antibacterial drugs. In addition, our dataset included informa-tion on whether the shortage was active or resolved, the reason for shortage, whether the product was an injectable product, whether the product was sole source (defined as a single-manufacturer product and determined by the investigator using FDA approval dates), and whether alternative drugs were available as a substitute.

The investigators further characterized each antibacterial drug in the data set to general pharmacologic class and whether the drug is broad spectrum or narrow spectrum in coverage (with broad-spectrum defined as treating both gram-positive and gram-negative infections). Further classification was con-ducted to determine special pathogen coverage (eg, for methi-cillin-resistantStaphylococcus aureus[MRSA] orPseudomonas aeruginosa). We excluded topical and ophthalmic antibacterial agents. For example, tobramycin powder shortages were exclud-ed because it is most commonly usexclud-ed in bone cement, although the product is labeled for injection. This is due to the increased

time to reconstitute for intravenous use. Although it is possible that some hospitals reconstitute the powder for intravenous use, including this information would inaccurately account for the ac-tual usage of systemic use. We also excluded antimycobacterial agents, such as rifampin, as this study’s focus was to evaluate shortages for antibiotics used to treat acute bacterial infections and not tuberculosis. Shortage duration was calculated for resolved shortages. Because this study did not include human subject data, it was considered exempt by our institutional review board. Data Analysis

Data were described using standard descriptive statistics. Uni-variate analysis found that drug shortage time (measured in days) was not normally distributed (right skewed). Therefore, we used the nonparametric 2-sample Wilcoxon rank-sum (Mann–Whitney) test to compare shortage times and describe average shortage time using median and interquartile range (IQR). Visual inspection of the drug shortages by month showed a distinct upward ( positive) trend from July 2007 to 2014. A segmented regression analysis of drugs shortages before and after July 2007 was performed to assess this trend. Time was coded as a continuous variable in 1-month intervals from 1 (January 2001) to 156 (December 2013). We excluded the first year of data collection (2001) from the regression equation to adjust for a time lag for drugs tofirst appear on the shortage list. For the purposes of analysis, drugs that were on shortage for >1 time period are counted as separate shortages. The total number of antibacterial drugs on shortage was regressed on a constant, a linear trend term (x), a dummy variable equaling zero prior to July 2007 and 1 thereafter, and a post–July 2007 trend term (x-78) as described by Wagner et al [12]. A Durbin-Watson test indicated presence of serial autocorrelation and was corrected for by using the Cochrane-Orcutt transformation [13]. APvalue of <.05 was considered significant. Data were collected in Microsoft Excel (Microsoft Corporation, Redmond, Washing-ton), and analyses were conducted using Stata 13.1 (College Station, Texas).

RESULTS

Over the course of the study period, 148 antibacterial drugs went on shortage. Of those, 112 were resolved by the end of the study period (31 December 2013), 10 were discontinued by the manufacturer, and 26 remained active. Figure1presents the trends in shortages over time.

The median number of new shortages per year was 10 (IQR, 7) and ranged from a low of 6 in 2004 to a high of 18 in 2011. The median duration of a resolved shortage was 232 days (IQR, 373.5 days), or approximately 7.5 months.

We examined resolved shortage times for drugs with and without injectable administration, drugs with and without

available alternatives, and broad- vs narrow-spectrum drugs. Drugs with injectable administration had a median shortage time of 250 days (IQR, 457 days) vs 129 days (IQR, 288 days) for drugs without injectable administration (P = .06).

Drugs without available alternatives had a median shortage of 149.5 days (IQR, 410 days) vs 262 days (IQR, 324 days) for drugs with an alternative (P = .10). Drugs classified as broad-spectrum had a median shortage of 232 days (IQR, 367 days) vs 199.5 days (IQR, 366 days) for narrow-spectrum drugs (P = .74).

Over the study period, an average of 14.2 drugs were on shortage per month (95% CI, 12.8–15.5). The number of drugs on shortage in a given month increased over time (Figure1). On average, there were 9.7 drugs on shortage per month prior to July 2007 vs 17.9 from July 2007 through De-cember 2013 (P< .001). From July 2007 to December 2013, the rate of increase was approximately 0.35 additional drugs every month (95% CI, .22–.49) and was significantly different from the trend prior to July 2007 (P< .001).

Thirty-two drugs (22%) experienced multiple shortages over the study period. The most commonly reported shortages were meropenem (7); cefotetan and piperacillin-tazobactam (5 each); and azithromycin, aztreonam, and nafcillin (4 each) (Table1). Cephalosporins were the most commonly reported drug class shortage, with 27 reported shortages for a total of 446 months over the study period. There were 11 aminoglycoside shortages for 284 months, 22 penicillin shortages for 229 months, and 11 penicillin/β-lactam inhibitor shortages for 178 months over the study period (Figure2).

One hundred twenty of the 138 drugs (81%) were broad-spectrum antibacterial agents. Trends are displayed in Figure3. Sixty-eight of the shortages (46%) involved drugs used to treat high-risk pathogens, including 2 forClostridium difficile, 1 for carbapenem-resistant Enterobacteriaceae (CRE), 1 for leprosy, 15 for MRSA, and 32 forPseudomonas aeruginosa.

Seven of the drugs treat extended-spectrum β-lactamase– producing organisms (ESBLs), and 4 treat human immu-nodeficiency virus–related opportunistic infections. Two drug formulations were specifically used to treat pediatric patients. The most common reason for shortage included manufacturing problems, with 35 reported shortages and 19 resulting from in-sufficient supply–demand ratio (Table2)

DISCUSSION

There were severalfindings of concern in this study. First, there were a substantial number of drug shortages from 2001 to 2013, Table 1. Drugs With Multiple Shortages: Both Active and Resolved, 2001–2013

Drug Name

No. of

Shortages Total Days

Cefotetan 5 2141 Aztreonam 4 1990 Piperacillin-tazobactam 5 ₁₈₅₈ Kanamycin injection 3 1682 Cefotaxime injection 3 ₁₅₄₂ Doxycycline injection 3 1515 Erythromycin lactobionate 2 1245

Gentamicin injection, pediatric strength 2 1173

Meropenem 7 ₁₁₁₄ Ampicillin-sulbactam _{3 1052} Minocycline injection 2 944 Piperacillin sodium 2 942 Cefepime injection 3 801 Nafcillin 4 763 Azithromycin injection 4 ₅₉₉ Gentamicin injection 2 ₅₇₄ Cephalexin suspension 2 ₅₃₅ Imipenem-cilastatin 2 502 Ciprofloxacin 500 mg tablets 3 ₄₄₂ Ceftizoxime 2 405 Clindamycin injection 2 ₃₉₄ Vancomycin 3 ₃₇₃ Cefpodoxime 2 ₃₄₄ Colistimethate 2 255 Paromomycin (Humatin) 3 192 Cefuroxime injection 2 ₁₈₅ Sulfamethoxazole-trimethoprim injection 3 ₁₆₅ Methenamine 2 133 Dalfopristin/quinupristin (Synercid) 2 125 Ticarcillin-clavulanate 3 ₁₂₃ Ceftazidime 2 ₆₉ Spectinomycin 2 0

Includes both active and resolved shortages. Values in boldface indicate that the drug is currently experiencing a shortage.

Source: University of Utah Drug Information Service database. Figure 1. Trends in drug shortages, 2001–2013.

with concerning trends over time. These shortages are often of very long duration, with a mean of nearly 9 months. Second, these shortages often impact clinicians’ability to treat infec-tions, including multidrug-resistant pathogens for which there is a limited selection of effective antibiotics. Finally, we found that a substantial proportion of shortages (61 of 148) were for unknown reasons, suggesting a need for improved tracking and reporting.

A high proportion of recent drug shortages involved broad-spectrum agents, injectable drugs, medications with no alternative sources, or those used on pathogens with limited alternative treat-ment options or in pediatric patients. Although we were unable to determine the precise cause for the dramatic rise in antibacterial shortages after 2007, the shortage increase coincides with the downturn in the US economy, with the most common reasons for shortages being business related. Since 2011, drug shortages appear to have stabilized, but still remain high.

The upward trend in shortages of broad-spectrum agents is driving the total increase in shortages. We found especially Figure 2. Drug shortages by drug class, 2001–2013.

Figure 3. Trends in broad-spectrum vs narrow-spectrum shortages, 2001–2013.

Table 2. Common Reasons for Drug Shortages, 2001–2013

Shortage Reason No. of Drugs

General manufacturing/manufacturing delays or problems

35

Supply and demand 19

Raw material shortage 13

Discontinued 10

Regulatory issues/regulatory problems 5

Regulatory (import ban) 3

Natural disaster 1

Business decision 1

Unknown 61

Total 148

concerning the high shortage rates in antibacterial agents used to treat highly drug-resistant pathogens, including MRSA, CRE, andPseudomonas. One hospital found that shortages occurred in 7 antimicrobials over a 3-month period, with 87% being ge-neric products [14]. Treatment options for CREs are likely to become increasingly limited [15,16].

Many shortages involve gold-standard therapies. For exam-ple, aztreonam treats life-threatening infections caused by gram-negative bacteria in patients allergic to penicillin and tri-methoprim-sulfamethoxazole (TMP-SMX) for treatment of

Pneumocystispneumonia (PCP) [2]. TMP-SMX shortages are compounded because it is produced by a single manufacturer. Seventy-three drugs (49%) were listed as having no alternative production source. Use of therapeutic alternatives include lower efficacy rates and higher toxicity [1]. The shortage of TMP-SMX necessitating treatment with alternatives, such as clindamycin and primaquine, has led to delayed care, as well as refractory cases of PCP [13].

Implications for the clinical management of patients with crit-ical infectious diseases in the setting of rising trends in drug resis-tance have been studied previously. A survey study of >600 infectious disease physicians found that 78% reported having to modify antibacterial choices due to shortages. In that study, clini-cians reported the antimicrobials most commonly unavailable or in short supply as TMP-SMX (65%), amikacin (58%), and aztreo-nam (31%), which is supported by ourfindings. More than half felt that the shortage had a negative impact on patient outcomes, requiring the use of less effective, more toxic, or more costly alter-natives. Most providers reported they learned of shortages after at-tempting to prescribe the medication through their pharmacy [5]. Alternatives may not be as familiar to clinicians and thus may lead to medication errors and adverse outcomes [1]. In a survey of the Institute for Safe Medication Practices conducted in 2010, clinicians reported adverse outcomes due to drug shortages in 20% of cases [17], as well as substantial resource utilization when developing an action plan (82%) and a lack of suitable al-ternatives (70%). Shortages may also lead to internal hoarding of medications, posing ethical challenges. Potential patient harm due to shortages includes medication overdoses, life-threatening side effects, cross-contamination, and even death [18]. For example, several patients withPseudomonasinfection only and amikacin sensitivity experienced morbidity or mortal-ity when the drug could not be provided [13–15].

There are several ways to potentially mitigate the clinical im-pact of antibacterial drug shortages. One important strategy is through improved communication with providers. In half of cases, physicians and nurses learn about drug shortages from pharmacy staff, often when the pharmacy is unable to dispense the medication [13]. Institutions should consider prospectively tracking potential shortages and making recommendations for safe and appropriate use of therapeutic alternatives [19].

Strategies should also include keeping local inventory of critical antibacterial agents, anticipating the need for alternatives, and creating contingency plans. A multidisciplinary antimicrobial stewardship plan should include a stewardship pharmacist to help guide these efforts. One strategy would be guideline devel-opment, ensuring narrow coverage based on culture results, en-couraging intravenous to oral conversion, ensuring ethical distribution to patients with the greatest need, educating clini-cians, and developing protocols for when alternatives must be used, so as to prevent errors [15]. An interdisciplinary drug shortage task force can include physicians, pharmacists, nurses, ethics members, and patient representatives [20]. Optimization of utilization of the available supply and guidance regarding al-ternative agents are imperative, with guidelines available [17].

Although more than a dozen new antimicrobial agents have recently been developed against antibiotic-resistant pathogens, few act on new targets; thus, it is imperative to address the lack of agents, compounded by drug shortages, for these classes [21]. Many times, manufacturers do not report shortages until supplies are very low, and policy makers should advocate for mandatory earlier reporting of shortages.

Limitations included the inability to account for all possible shortages, lack of information on geographic disparities in short-ages, and evidence of impact on clinical outcomes. Drug shortages may impact individual healthcare facilities at different times and with differing effects on clinical care. For example, not all facilities stock all drugs or all formulations. Small or rural facilities may have more difficulty accessing products that can only be pur-chased directly from the manufacturer. Inventory and delivery methods vary widely in the drug distribution chain. Furthermore, it is somewhat difficult to distinguish if the drug was completely unavailable, discontinued, or available but in short supply, so we cannot describe the severity of the shortage. We were also unable to measure the time and effort hospital staff spends attempting to mitigate shortages. Nonetheless, the database currently used rep-resents the most comprehensive data source on trends and rea-sons for antibacterial and other drug shortages.

In summary, there were 148 reported antibacterial shortages, with 26 drugs still actively on shortage as of 31 December 2013, with nearly half reported as having no alternative production source. Given that 46% are used in the treatment of high-risk pathogens, such as MRSA, CRE, andPseudomonas, it is imper-ative to develop a comprehensive strategy to mitigate the impli-cations of shortages on clinical practice and patient outcomes, including improved communication, alternative algorithms, and antimicrobial stewardship policies.

Notes

Author contributions. F. Q. was responsible for the literature review, analysis of data, results, editing, and abstract. M. M.-A. was responsible for study design, data collection, data interpretation, and writing. E. R. F.

was responsible for data collection, data interpretation, and writing. K. L. H. was responsible for study design, data collection, and data interpretation. J. M. P. was responsible for study design, data interpre-tation, and writing. M. S. Z. was responsible for data analysis,figures, and editing. L. M. was responsible for study design, data analysis, data interpre-tation, literature review, writing, and editing. The corresponding author had full access to all study data and hadfinal responsibility for the decision to submit for publication.

Potential conflicts of interest. All authors: No potential conflicts of interest.

All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the con-tent of the manuscript have been disclosed.

References

1. Krisl JC, Fortier CR, Taber DJ. Disruptions in the supply of medications used in transplantation: implications and management strategies for the transplant clinician. Am J Transplant2013; 13:20–30.

2. US Food and Drug Administration. Center for Drug Evaluation and Research. Office of New Drugs, Drug Shortage Management,2012. Manual of policies and procedures. Available at: http://www.fda. gov/downloads/AboutFDA/CentersOffices/CDER/ManualofPolicies Procedures/ucm079936.pdf. Accessed 28 August 2014.

3. Woodcock J, Wosinska M. Economic and technological drivers of ge-neric sterile injectable drug shortages. Clin Pharmacol Ther2013; 93:170–6.

4. Aitken M. Prescription drug shortages: examining a public health con-cern and potential solutions. Parsippany, NJ: IMS Institute for Health-care Informatics,2011.

5. Gundlapalli AV, Beekmann SE, Graham DR, Polgreen PM; Infectious Diseases Society of America’s Emerging Infections Network. Perspec-tives and concerns regarding antimicrobial agent shortages among in-fectious disease specialists. Diagn Microbiol Infect Dis2013; 75:256–9. 6. Gould IM. Coping with antibiotic resistance: the impending crisis. Int J

Antimicrob Agents2010; 36(Suppl 3):S1–2.

7. Spellberg B, Shlaes D. Prioritized current unmet needs for antibacterial therapies. Clin Pharmacol Ther2014; 96:151–3.

8. Falagas ME, Tansarli GS, Karageorgopoulos DE, Vardakas KZ. Deaths attributable to carbapenem-resistant Enterobacteriaceae infections. Emerg Infect Dis2014; 20:1170–5.

9. Drug Shortages: Public Health Threat Continues, Despite Efforts to Help Ensure Product Availability. US Government Accountability

Of-fice. Available at:http://www.gao.gov/products/GAO-14-194. Accessed 22 July 2014.

10. American Society of Health-System Pharmacists. ASHP guidelines on managing drug product shortages in hospitals and health systems. Am J Health Syst Pharm2009; 66:1399–406.

11. Fox ER, Tyler LS. Managing drug shortages: seven years’experience at one health system. Am J Health Syst Pharm2003; 60:245–53. 12. Wagner AK, Soumerai FZ, Zhang F, Ross-Degnan D. Segmented

regres-sion analysis of interrupted time series studies in medication use research. J Clin Pharm Ther2002; 27:299–309.

13. Cochrane D, Orcutt GH. Application of least squares regression to re-lationships containing auto-correlated error terms. J Am Stat Assoc

1949; 44:32–61.

14. Griffith MM, Pentoney Z, Scheetz MH. Antimicrobial drug shortages: a crisis amidst the epidemic and the need for antimicrobial stewardship efforts to lessen the effects. Pharmacotherapy2012; 32:665–7. 15. Wailan AM, Paterson DL. The spread and acquisition of NDM-1:

a multifactorial problem. Expert Rev Anti Infect Ther2014; 12: 91–115.

16. Biedenbach DJ, Bouchillon SK, Hoban DJ, et al. Antimicrobial susceptibility and extended-spectrum beta-lactamase rates in aero-bic gram-negative bacteria causing intra-abdominal infections in Vietnam: report from the Study for Monitoring Antimicrobial Resis-tance Trends (SMART 2009–2011). Diagn Microbiol Infect Dis2014; 79:463–7.

17. Institute for Safe Medication Practices. Drug shortages: national survey reveals high level of frustration, low level of safety. Horsham, PA: Insti-tute for Safe Medication Practices,2010; 15.

18. Institute for Safe Medication Practices. A shortage of everything except errors: harm associated with drug shortages. Horsham, PA: Institute for Safe Medication Practices,2012; 17.

19. Griffith MM, Patel JA, Sutton SH, et al. Prospective approach to man-aging antimicrobial drug shortages. Infect Control Hosp Epidemiol

2012; 33:745–52.

20. Valgas J, Singer EA, Berry SR, Kimryn Rathmell W. Ethical challenges: managing oncology drug shortages. J Oncol Pract2013; 9:e21–3. 21. Freire-Moran L, Aronsson B, Manz C, et al. Critical shortage of new

an-tibiotics in development against multidrug-resistant bacteria: time to react is now. Drug Resist Updat2011; 14:118–24.